Identifying Risk Factors for Depression and Anxiety Among Rural Children in China: A Machine Learning Approach

To identify potential risk factors for depression and anxiety among rural Chinese children, this study tested and compared the performance of six machine learning algorithms: Logistic Regression, Naive Bayes, Decision Tree, Random Forest, K-Nearest Neighbors(KNN), and Light Gradient Boosting Machine (LightGBM), based on a psychological health dataset for rural children in China. The results indicated that Random Forest and LightGBM outperformed the other algorithms regarding predictive accuracy and precision. Specifically, these two algorithms demonstrated a significant advantage over the other four machine learning methods. In further analysis, we evaluated the predictive power of 22 potential factors associated with the risk of depression and anxiety-such as peer relationships, paternal involvement, children's behavioral issues, and parental relationships-using the feature importance of both Random Forest and LightGBM. The results revealed that this combined model effectively distinguished between high-risk and low-risk factors for depression and anxiety among rural Chinese children. This highlights the model's significant application value in the prevention of depression and anxiety in this population.